236 lines
6.8 KiB
Rust
Raw Normal View History

feat: Phase 5.3 - Multi-Agent Learning Infrastructure Implement intelligent agent learning from Knowledge Graph execution history with per-task-type expertise tracking, recency bias, and learning curves. ## Phase 5.3 Implementation ### Learning Infrastructure (✅ Complete) - LearningProfileService with per-task-type expertise metrics - TaskTypeExpertise model tracking success_rate, confidence, learning curves - Recency bias weighting: recent 7 days weighted 3x higher (exponential decay) - Confidence scoring prevents overfitting: min(1.0, executions / 20) - Learning curves computed from daily execution windows ### Agent Scoring Service (✅ Complete) - Unified AgentScore combining SwarmCoordinator + learning profiles - Scoring formula: 0.3*base + 0.5*expertise + 0.2*confidence - Rank agents by combined score for intelligent assignment - Support for recency-biased scoring (recent_success_rate) - Methods: rank_agents, select_best, rank_agents_with_recency ### KG Integration (✅ Complete) - KGPersistence::get_executions_for_task_type() - query by agent + task type - KGPersistence::get_agent_executions() - all executions for agent - Coordinator::load_learning_profile_from_kg() - core KG→Learning integration - Coordinator::load_all_learning_profiles() - batch load for multiple agents - Convert PersistedExecution → ExecutionData for learning calculations ### Agent Assignment Integration (✅ Complete) - AgentCoordinator uses learning profiles for task assignment - extract_task_type() infers task type from title/description - assign_task() scores candidates using AgentScoringService - Fallback to load-based selection if no learning data available - Learning profiles stored in coordinator.learning_profiles RwLock ### Profile Adapter Enhancements (✅ Complete) - create_learning_profile() - initialize empty profiles - add_task_type_expertise() - set task-type expertise - update_profile_with_learning() - update swarm profiles from learning ## Files Modified ### vapora-knowledge-graph/src/persistence.rs (+30 lines) - get_executions_for_task_type(agent_id, task_type, limit) - get_agent_executions(agent_id, limit) ### vapora-agents/src/coordinator.rs (+100 lines) - load_learning_profile_from_kg() - core KG integration method - load_all_learning_profiles() - batch loading for agents - assign_task() already uses learning-based scoring via AgentScoringService ### Existing Complete Implementation - vapora-knowledge-graph/src/learning.rs - calculation functions - vapora-agents/src/learning_profile.rs - data structures and expertise - vapora-agents/src/scoring.rs - unified scoring service - vapora-agents/src/profile_adapter.rs - adapter methods ## Tests Passing - learning_profile: 7 tests ✅ - scoring: 5 tests ✅ - profile_adapter: 6 tests ✅ - coordinator: learning-specific tests ✅ ## Data Flow 1. Task arrives → AgentCoordinator::assign_task() 2. Extract task_type from description 3. Query KG for task-type executions (load_learning_profile_from_kg) 4. Calculate expertise with recency bias 5. Score candidates (SwarmCoordinator + learning) 6. Assign to top-scored agent 7. Execution result → KG → Update learning profiles ## Key Design Decisions ✅ Recency bias: 7-day half-life with 3x weight for recent performance ✅ Confidence scoring: min(1.0, total_executions / 20) prevents overfitting ✅ Hierarchical scoring: 30% base load, 50% expertise, 20% confidence ✅ KG query limit: 100 recent executions per task-type for performance ✅ Async loading: load_learning_profile_from_kg supports concurrent loads ## Next: Phase 5.4 - Cost Optimization Ready to implement budget enforcement and cost-aware provider selection.
2026-01-11 13:03:53 +00:00
// vapora-backend: Workflow state machine
// Phase 3: State management for workflow lifecycle
use chrono::{DateTime, Utc};
use serde::{Deserialize, Serialize};
#[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
pub enum WorkflowStatus {
Created,
Planning,
InProgress,
Blocked,
Completed,
Failed,
RolledBack,
}
#[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
pub enum StepStatus {
Pending,
Running,
Completed,
Failed,
Skipped,
Blocked,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Workflow {
pub id: String,
pub title: String,
pub status: WorkflowStatus,
pub phases: Vec<Phase>,
pub created_at: DateTime<Utc>,
pub started_at: Option<DateTime<Utc>>,
pub completed_at: Option<DateTime<Utc>>,
pub estimated_completion: Option<DateTime<Utc>>,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Phase {
pub id: String,
pub name: String,
pub status: StepStatus,
pub steps: Vec<WorkflowStep>,
pub parallel: bool,
pub estimated_hours: f32,
}
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct WorkflowStep {
pub id: String,
pub name: String,
pub agent_role: String,
pub status: StepStatus,
pub depends_on: Vec<String>,
pub can_parallelize: bool,
pub started_at: Option<DateTime<Utc>>,
pub completed_at: Option<DateTime<Utc>>,
pub result: Option<String>,
pub error: Option<String>,
}
impl Default for WorkflowStep {
fn default() -> Self {
Self {
id: String::new(),
name: String::new(),
agent_role: String::new(),
status: StepStatus::Pending,
depends_on: Vec::new(),
can_parallelize: false,
started_at: None,
completed_at: None,
result: None,
error: None,
}
}
}
impl Workflow {
/// Create a new workflow
pub fn new(id: String, title: String, phases: Vec<Phase>) -> Self {
Self {
id,
title,
status: WorkflowStatus::Created,
phases,
created_at: Utc::now(),
started_at: None,
completed_at: None,
estimated_completion: None,
}
}
/// Check if transition is allowed
pub fn can_transition(&self, to: &WorkflowStatus) -> bool {
match (&self.status, to) {
(WorkflowStatus::Created, WorkflowStatus::Planning) => true,
(WorkflowStatus::Planning, WorkflowStatus::InProgress) => true,
(WorkflowStatus::InProgress, WorkflowStatus::Completed) => true,
(WorkflowStatus::InProgress, WorkflowStatus::Failed) => true,
(WorkflowStatus::InProgress, WorkflowStatus::Blocked) => true,
(WorkflowStatus::Blocked, WorkflowStatus::InProgress) => true,
(WorkflowStatus::Failed, WorkflowStatus::RolledBack) => true,
_ => false,
}
}
/// Transition to new state
pub fn transition(&mut self, to: WorkflowStatus) -> Result<(), String> {
if !self.can_transition(&to) {
return Err(format!(
"Cannot transition from {:?} to {:?}",
self.status, to
));
}
match &to {
WorkflowStatus::InProgress => {
self.started_at = Some(Utc::now());
}
WorkflowStatus::Completed | WorkflowStatus::Failed | WorkflowStatus::RolledBack => {
self.completed_at = Some(Utc::now());
}
_ => {}
}
self.status = to;
Ok(())
}
/// Check if all steps are completed
pub fn all_steps_completed(&self) -> bool {
self.phases.iter().all(|p| {
p.steps
.iter()
.all(|s| matches!(s.status, StepStatus::Completed | StepStatus::Skipped))
})
}
/// Check if any step has failed
pub fn any_step_failed(&self) -> bool {
self.phases
.iter()
.any(|p| p.steps.iter().any(|s| matches!(s.status, StepStatus::Failed)))
}
/// Get workflow progress percentage
pub fn progress_percent(&self) -> u32 {
let total_steps: usize = self.phases.iter().map(|p| p.steps.len()).sum();
if total_steps == 0 {
return 0;
}
let completed_steps: usize = self
.phases
.iter()
.flat_map(|p| &p.steps)
.filter(|s| matches!(s.status, StepStatus::Completed | StepStatus::Skipped))
.count();
((completed_steps as f64 / total_steps as f64) * 100.0) as u32
}
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_workflow_creation() {
let workflow = Workflow::new("wf-1".to_string(), "Test Workflow".to_string(), vec![]);
assert_eq!(workflow.id, "wf-1");
assert_eq!(workflow.status, WorkflowStatus::Created);
assert!(workflow.started_at.is_none());
}
#[test]
fn test_valid_transitions() {
let mut workflow = Workflow::new("wf-1".to_string(), "Test".to_string(), vec![]);
assert!(workflow.transition(WorkflowStatus::Planning).is_ok());
assert_eq!(workflow.status, WorkflowStatus::Planning);
assert!(workflow.transition(WorkflowStatus::InProgress).is_ok());
assert_eq!(workflow.status, WorkflowStatus::InProgress);
assert!(workflow.started_at.is_some());
assert!(workflow.transition(WorkflowStatus::Completed).is_ok());
assert_eq!(workflow.status, WorkflowStatus::Completed);
assert!(workflow.completed_at.is_some());
}
#[test]
fn test_invalid_transition() {
let mut workflow = Workflow::new("wf-1".to_string(), "Test".to_string(), vec![]);
let result = workflow.transition(WorkflowStatus::Completed);
assert!(result.is_err());
}
#[test]
fn test_progress_calculation() {
let mut workflow = Workflow::new(
"wf-1".to_string(),
"Test".to_string(),
vec![Phase {
id: "p1".to_string(),
name: "Phase 1".to_string(),
status: StepStatus::Running,
parallel: false,
estimated_hours: 2.0,
steps: vec![
WorkflowStep {
id: "s1".to_string(),
status: StepStatus::Completed,
..Default::default()
},
WorkflowStep {
id: "s2".to_string(),
status: StepStatus::Running,
..Default::default()
},
],
}],
);
assert_eq!(workflow.progress_percent(), 50);
workflow.phases[0].steps[1].status = StepStatus::Completed;
assert_eq!(workflow.progress_percent(), 100);
}
}